Re: svn commit: rev 1511 - trunk/subversion/include trunk/subversion/libsvn_ra_local trunk/subversion/libsvn_client trunk/subversion/libsvn_ra_dav

Greg Stein <gstein@lyra.org> writes:

> > - /* Retrieve a "working copy" property for an item relative to the
> > - "root" of the session (defined by REPOS_URL to the open() vtable
> > - function call).
> > -
> > - The baton passed to the get_wcprop callback is the CALLBACK_BATON.
> > -
> > - If it quite legal to set this field to NULL if the RA user cannot
> > - support WC properties, or they are not defined/applicable for the
> > - current session.
> > -
> > - ### we might have a problem if the RA layer ever wants a property
> > - ### that corresponds to a different revision of the file than
> > - ### what is in the WC. we'll cross that bridge one day...
> > - */
> > + /* Fetch working copy properties. */
> > svn_ra_get_wc_prop_func_t get_wc_prop;
>
> Woah! Why did you torch this comment? That removed all knowledge about
> being legal to set that field to NULL.

Actually, I should have made a blanket comment in that callback baton
that states that it's okay to have get_wc_prop, set_wc_prop, or
close_commit set to NULL. Good catch.

> This also removed the warning about retrieving props for a revision that
> doesn't occur in the WC.

I thought I simply moved that comment elsewhere. I'll re-check, and
if not, will re-add.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Mar 14 20:58:52 2002